JPH0895587A - Text-to-speech method - Google Patents

Abstract

(57) [Summary] [Purpose] To be able to perform speech synthesis for all text data even if English, other languages, dialects, etc. are mixed in Japanese sentences. [Structure] Dictionary information set corresponding to character modification is added to a character code to generate text data, and the text data is analyzed. The text-to-speech synthesizer 13
A dictionary corresponding to each language is selected on the basis of the dictionary information of the analyzed text data, and speech synthesis is performed by referring to the selected dictionary. Even if various languages and dialects are mixed in one sentence, it is possible to perform speech synthesis on all text data. Further, since the dictionary information can be set in correspondence with the character modification, no special control is required, and voice synthesis can be performed using a normal word processor, personal computer or the like.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a text-to-speech method.

[0002]

2. Description of the Related Art Conventionally, a text-to-speech synthesizer receives text data of a kanji-kana mixed sentence from a higher-level device such as a word processor or a personal computer, analyzes the text data, and synthesizes a voice. Therefore, the text-to-speech synthesizer has a built-in accent dictionary, user dictionary, etc., and can synthesize speech by referring to these dictionaries. An input interface (I / F) is arranged between the host device and the text-to-speech synthesizer, and the text data is input to the text-speech synthesizer via the input interface.

In this case, a Centronics standard interface for a printer is used as the input interface.

[0004]

However, in the conventional text-to-speech synthesizer, the accent dictionaries are all dictionaries corresponding to only one language, and the speech synthesis is performed only on the text data represented by the language. You can do it. Therefore, for example, in a text-to-speech synthesizer with a built-in dictionary corresponding to Japanese, if English, another language, dialect, etc. are mixed in a Japanese sentence, the English, other language,
Speech synthesis cannot be performed on text data such as dialects.

The present invention solves the above-mentioned problems of the conventional text-to-speech synthesizer, and speech-synthesizes all text data even if English, other languages, dialects, etc. are mixed in a Japanese sentence. It is an object of the present invention to provide a text-to-speech method capable of performing.

[0006]

Therefore, in the text-to-speech synthesis method of the present invention, the dictionary information set corresponding to the character modification is added to the character code to generate the text data, and the text data is converted. A dictionary corresponding to each language is selected based on the analyzed dictionary information of the analyzed text data, and speech synthesis is performed with reference to the selected dictionary.

[0007]

According to the present invention, as described above, in the text-to-speech synthesis method, text data is generated by adding the dictionary information set in advance in correspondence with the character modification to the character code. In the text-to-speech synthesizer, the text data in which the dictionary information is added to the character code is received from the host device, and the text data is analyzed. Further, the text-to-speech synthesizer has a dictionary corresponding to each language, and selects a dictionary corresponding to each language based on the dictionary information of the analyzed text data.

Then, the text-to-speech synthesizer refers to the selected dictionary to generate waveform data and synthesizes the speech.

[0009]

Embodiments of the present invention will now be described in detail with reference to the drawings. FIG. 1 is a block diagram of a text-to-speech synthesizer according to an embodiment of the present invention. In the drawing, 11 is a word processor, a host device such as a personal computer,
Reference numeral 12 is a dedicated printer driver arranged in the host device 11. Further, 13 is a text-to-speech synthesizer, and the text-to-speech synthesizer 13 receives text data from the higher-level device 11 via the printer interface 14.

The text-to-speech synthesizer 13 has a dictionary selection section 16, a dictionary section 17 and a speech synthesis section 18. The dictionary selection unit 16 is arranged in a printer interface processing unit (not shown) of the text-to-speech synthesizer 13 and is composed of software for selecting a dictionary to be used. The dictionary unit 17 includes dictionaries corresponding to respective languages, for example, a Japanese dictionary 21, an English dictionary 22, a French dictionary 23, a German dictionary 24, and a user dictionary 25. Then, the voice synthesizer 18 analyzes the text data sent via the interface 14, generates waveform data, and performs voice synthesis.

In this case, the text data created in the host device 11 for selecting the dictionary is subjected to character modification such as italics, underlining, and shading.
FIG. 2 is an explanatory diagram of text data in the embodiment of the present invention. In the figure, 31 is a screen of the higher-level device 11 (FIG. 1), and Japanese characters 32 and English characters 3 are included in the sentence.
3, French letters 34, German letters 35, etc. are mixed. In this case, in order to distinguish each language, by operating an input device (not shown), the Japanese character 32 is expressed by a normal typeface, and the English character 33 is italicized by French. In the case of the character 34, the underlined character modification is performed, and in the case of the German character 35, the shaded character modification is performed.

The characters 32 to 35 are converted into text data 38 by the dedicated printer driver 12 and output via the interface 14. In this case, the text data 38 includes the character code strings 39, 42, 48 representing the characters 32 to 35, the country designation control codes 40, 4
6, country designation code 41, 47, cancel command CA
N, image transfer control codes 44, 50, and image data 45, 5 for printing characters with character modification
It consists of 1. In this case, the country designated control code 40, 4
6 and country designation codes 41 and 47 constitute dictionary information.

The country designation control codes 40 and 46 indicate that the character code strings 39, 42 and 48 following the code do not correspond to the Japanese dictionary 21 or the user dictionary 25. Further, the country designation codes 41, 47 indicate the respective languages of the character code strings 39, 42, 48 that follow. The cancel command CAN indicates that the preceding country designation control codes 40, 46, country designation codes 41, 47, and character code strings 39, 42, 48 are canceled. Further, the image transfer control codes 44, 50
Indicates that the subsequent text data 38 is image data. The country-specific control codes 40, 46 also include the number of characters of the target character code strings 39, 42, 48 as a parameter.

Next, the text-to-speech synthesizer 1 having the above configuration
The operation of No. 3 will be described with reference to FIGS. First, the operator operates the input device of the host device 11,
Create a sentence in which each language such as Japanese, English, French, and German is mixed. At this time, for characters other than Japanese and characters registered in the user dictionary, preset character decoration is performed. In this embodiment, the English character 33
Is italicized, and French 34 is underlined, German 35 is
In the case of, the characters are shaded. However, the character decoration corresponding to each language can be freely selected.

Next, when a printing process, that is, a voice output process is performed after a sentence is created, the dedicated printer driver 12 incorporated in the higher-level device 11 makes Japanese characters for characters that are not character-modified. Character code string 39 is created, and for characters that have been character-modified, a character code string 42 in which country-specific control codes 40 and 46 and country-specific codes 41 and 47 are added according to the content of the character modification. ,
48 is generated and the text data 38 is generated.

Image data 45 and 51 for print output processing are also formed in the text data 38 so that print output processing can be performed in addition to the voice output processing. When the host device 11 is connected to the text-to-speech synthesizer 13, the text-to-speech synthesizer 13 causes the image data 45 in the text data 38,
Ignore 51.

On the other hand, when the upper device 11 is connected to a printer (not shown), the image data 45, 51
Becomes the print data. In this case, country-specific control code 4
0, 46, country designation code 41, 47 and character code string 4
Nos. 2 and 48 are ignored by the cancel command CAN that follows, and are not printed. On the other hand, when the text-to-speech synthesizer 13 receives the text data 38 from the higher-level device 11, the dictionary selection unit 16 causes the country designation control codes 40 and 46.
It is determined which language the character code strings 39, 42 and 48 of the text data 38 correspond to based on the country designation codes 41 and 47, and the Japanese dictionary 21 and the English dictionary 22 of the dictionary unit 17, The dictionary of the corresponding language is selected from the French dictionary 23, the German dictionary 24, and the user dictionary 25.

Next, the voice synthesizing unit 18 refers to the dictionary of the selected language, analyzes the character code strings 39, 42 and 48 of the text data 38, generates waveform data and performs voice synthesis. Then, the output unit (not shown) outputs the analog signal obtained by the voice synthesis at a predetermined level. In the present embodiment, the dictionary for each language is formed in the dictionary unit 17, but it is also possible to form a dialect dictionary such as Kansai dialect, Kyoto dialect, Tohoku dialect.

Next, the operation of the text-to-speech synthesizer 13 will be described with reference to FIGS. FIG. 3 is a flow chart showing the operation of the text-to-speech synthesis method in the embodiment of the present invention, and FIG. 4 is a flow chart showing the subroutine of the dictionary selection processing in the embodiment of the present invention. Step S1 Dictionary selection processing is performed based on the text data 38 (FIG. 2) sent from the higher-level device 11 (FIG. 1). In step S2, the text data 38 is analyzed and waveform data is generated. Step S3 The voice synthesizing unit 18 performs voice synthesizing based on the waveform data, and an output unit (not shown) generates an analog signal and outputs it at a predetermined level. Step S1-1 The sent text data 38 is analyzed, and it is determined whether the analyzed text data is the country designated control codes 40 and 46. If it is the country designated control code 40 or 46, the process proceeds to step S1-2. If it is not the country designated control code 40 or 46, the process proceeds to step S1-4. In step S1-2, the language is judged based on the country designation codes 41 and 47. Step S1-3 Select a dictionary for each language. Step S1-4: It is judged whether the analyzed text data is another control code. If it is another control code, the process proceeds to step S1-5. If it is not another control code, the process proceeds to step S1-6. In step S1-5, control code processing is performed. Step S1-6 Since the character code is known, the Japanese dictionary 21 is selected.

The present invention is not limited to the above-mentioned embodiments, but can be variously modified within the scope of the present invention, and these are not excluded from the scope of the present invention.

[0021]

As described above in detail, according to the present invention, in the text-to-speech synthesis method, the dictionary information set corresponding to the character modification is added to the character code to generate the text data, The text data is analyzed. Further, the text-to-speech synthesizer selects a dictionary corresponding to each language based on the dictionary information of the analyzed text data, and refers to the selected dictionary to perform speech synthesis.

Therefore, even if various languages, dialects and the like are mixed in one sentence, it is possible to perform speech synthesis on all text data. Further, since the dictionary information can be set in correspondence with the character modification, no special control is required, and voice synthesis can be performed using a normal word processor, personal computer or the like. Also,
It is also possible to connect the host device and the text-to-speech synthesizer via an interface for a printer.

[Brief description of drawings]

FIG. 1 is a block diagram of a text-to-speech synthesizer according to an embodiment of the present invention.

FIG. 2 is an explanatory diagram of text data according to the embodiment of this invention.

FIG. 3 is a flowchart showing the operation of the text-to-speech synthesis method in the embodiment of the present invention.

FIG. 4 is a flowchart showing a subroutine of dictionary selection processing in the embodiment of the present invention.

[Explanation of symbols]

 11 Upper Device 12 Dedicated Printer Driver 13 Text-to-Speech Synthesizer 16 Dictionary Selection Section 18 Speech-Synthesis Section 38 Text Data

Claims (1)

[Claims] 1. (a) The dictionary information set corresponding to the character modification is added to the character code to generate text data, and (b) the text data is analyzed, and the dictionary information of the analyzed text data. A text-to-speech synthesis method, characterized in that a dictionary corresponding to each language is selected based on, and (c) speech synthesis is performed with reference to the selected dictionary.